[NeurIPS 2020] Official code for the paper "DeepSVG: A Hierarchical Generative Network for Vector Graphics Animation". Includes a PyTorch library for deep learning with SVG data.
How would one go about extending the library to include a differentiable SVG.draw() function? Ideally, this would allow a neural network to flow gradients from the output SVG image to the SVG code.
How would one go about extending the library to include a differentiable
SVG.draw()function? Ideally, this would allow a neural network to flow gradients from the output SVG image to the SVG code.Thanks!